DESCRIPTION: The investigator proposes to extend group sequential methods for early stopping of clinical trials and epidemiologic studies. The specific aims of the proposal are: 1) to develop group-sequential methods for clinical trials in which some portion of the patients are considered cured of the disease, 2) to develop group-sequential methods for clinical trials with repeated measures, e.g., longitudinal studies, potentially with missing data, 3) to develop group-sequential methods for epidemiologic studies such as retrospective case-control studies and prospective cohort studies, and 4) to develop "work-arounds" to the problem of 'overrunning' and 'underrunning' the maximum information in long-term clinical trials. Development of practical guidelines for implementation of the proposed methods will be emphasized, and data from clinical trials will be utilized in investigating the practical aspects of the proposed methods. The group sequential methods rely primarily on a design paradigm using the Lan-DeMets (1983) type I error spending function approach, which can be implemented in a standard way when an 'independent increments' structure of sequentially computed test statistics is established.